#vertexai — Public Fediverse posts

生成AIによる処理結果をTP/FP/TN/FN×信頼区間でテストする
https://qiita.com/take-yoda/items/4f561e8ad7c069c71001?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Python #初心者向け #pytest #VertexAI #生成AI

生成AIによる処理結果をTP/FP/TN/FN×信頼区間でテストする
https://qiita.com/take-yoda/items/4f561e8ad7c069c71001?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Python #初心者向け #pytest #VertexAI #生成AI

Gemini Enterprise に ADK エージェントを登録して、スケジュール実行を試みたら詰まった話
https://qiita.com/re-watanabe/items/4aac7ebd4ad493be92b3?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Python #ADK #GoogleCloud #VertexAI #GeminiEnterprise

Gemini Enterprise に ADK エージェントを登録して、スケジュール実行を試みたら詰まった話
https://qiita.com/re-watanabe/items/4aac7ebd4ad493be92b3?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Python #ADK #GoogleCloud #VertexAI #GeminiEnterprise

Google 検索グラウンディングを検証してみた 〜画像検索・Maps 対応と課金の注意点〜
https://qiita.com/shun-aoki/items/ee2b63d170008464f9ac?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Gemini #GoogleCloud #VertexAI #生成AI #AgentPlatform

Google 検索グラウンディングを検証してみた 〜画像検索・Maps 対応と課金の注意点〜
https://qiita.com/shun-aoki/items/ee2b63d170008464f9ac?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Gemini #GoogleCloud #VertexAI #生成AI #AgentPlatform

最新AIエージェント機能16連発！徹底解説＆コンソール画面で検証してみた
https://qiita.com/harunatotsuka/items/cdb79a2578d2d50e9039?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Agent #GoogleCloudNext #GoogleCloud #VertexAI #GeminiEnterprise

最新AIエージェント機能16連発！徹底解説＆コンソール画面で検証してみた
https://qiita.com/harunatotsuka/items/cdb79a2578d2d50e9039?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Agent #GoogleCloudNext #GoogleCloud #VertexAI #GeminiEnterprise

AIの嘘を見抜く？Gemini の検索＆画像グラウンディングを実際に試してわかったこと
https://qiita.com/ta-ito0611/items/646d3d1efb535737416d?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Python #AI #Gemini #GoogleCloud #VertexAI

AIの嘘を見抜く？Gemini の検索＆画像グラウンディングを実際に試してわかったこと
https://qiita.com/ta-ito0611/items/646d3d1efb535737416d?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Python #AI #Gemini #GoogleCloud #VertexAI

GeminiでText-to-SQLの精度改善を実験してみた
https://qiita.com/take-yoda/items/6392ad3b5921a1d975b6?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #SQL #RDBMS #VertexAI #生成AI #TextToSQL

GeminiでText-to-SQLの精度改善を実験してみた
https://qiita.com/take-yoda/items/6392ad3b5921a1d975b6?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #SQL #RDBMS #VertexAI #生成AI #TextToSQL

⚠️ CRITICAL: Vertex AI Python SDK vulnerability enables cross-tenant RCE via bucket squatting during model uploads. No CVE or patch yet. Restrict bucket/model permissions & stay alert for updates. Unit 42 analysis: https://radar.offseq.com/threat/pickle-in-the-middle-hijacking-vertex-ai-model-upl-352aef09 #OffSeq #VertexAI #infosec #cloud

⚠️ CRITICAL: Vertex AI Python SDK vulnerability enables cross-tenant RCE via bucket squatting during model uploads. No CVE or patch yet. Restrict bucket/model permissions & stay alert for updates. Unit 42 analysis: https://radar.offseq.com/threat/pickle-in-the-middle-hijacking-vertex-ai-model-upl-352aef09 #OffSeq #VertexAI #infosec #cloud

⚠️ CRITICAL: Vertex AI Python SDK vulnerability enables cross-tenant RCE via bucket squatting during model uploads. No CVE or patch yet. Restrict bucket/model permissions & stay alert for updates. Unit 42 analysis: https://radar.offseq.com/threat/pickle-in-the-middle-hijacking-vertex-ai-model-upl-352aef09 #OffSeq #VertexAI #infosec #cloud

⚠️ CRITICAL: Vertex AI Python SDK vulnerability enables cross-tenant RCE via bucket squatting during model uploads. No CVE or patch yet. Restrict bucket/model permissions & stay alert for updates. Unit 42 analysis: https://radar.offseq.com/threat/pickle-in-the-middle-hijacking-vertex-ai-model-upl-352aef09 #OffSeq #VertexAI #infosec #cloud

Google Cloud's Vertex AI: A Hub for Generative AI Development Navigates API Access and Integration

Google Vertex AI users get 403 errors. Learn how to fix API key and service account permissions for Gemini and other AI models.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM

https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

Google Cloud's Vertex AI: A Hub for Generative AI Development Navigates API Access and Integration

Google Vertex AI users get 403 errors. Learn how to fix API key and service account permissions for Gemini and other AI models.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM

https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

Google Cloud's Vertex AI: A Hub for Generative AI Development Navigates API Access and Integration

Google Vertex AI users get 403 errors. Learn how to fix API key and service account permissions for Gemini and other AI models.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM

https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

Google Cloud's Vertex AI: A Hub for Generative AI Development Navigates API Access and Integration

Google Vertex AI users get 403 errors. Learn how to fix API key and service account permissions for Gemini and other AI models.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM

https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

Google Vertex AI is making it easier to build AI, but many users are hitting a wall with API key errors. This is a common problem for developers.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM
https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

Google Vertex AI is making it easier to build AI, but many users are hitting a wall with API key errors. This is a common problem for developers.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM
https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

Google Vertex AI is making it easier to build AI, but many users are hitting a wall with API key errors. This is a common problem for developers.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM
https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

Google Vertex AI is making it easier to build AI, but many users are hitting a wall with API key errors. This is a common problem for developers.

#VertexAI, #GoogleCloud, #APIKeys, #GenerativeAI, #IAM
https://newsletter.tf/google-vertex-ai-api-key-permission-errors/

→ Mindlid: 20% lift in top-1 recall for wellness app

🔗 Available via #GeminiAPI & #VertexAI
Integrates with #LangChain, #LlamaIndex, #Haystack, #Weaviate, #Qdrant, #ChromaDB & Vector Search
🧪 Colab notebooks for Gemini API & Vertex AI ready to use

🌐
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2-generally-available/

→ Mindlid: 20% lift in top-1 recall for wellness app

🔗 Available via #GeminiAPI & #VertexAI
Integrates with #LangChain, #LlamaIndex, #Haystack, #Weaviate, #Qdrant, #ChromaDB & Vector Search
🧪 Colab notebooks for Gemini API & Vertex AI ready to use

🌐
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2-generally-available/

→ Mindlid: 20% lift in top-1 recall for wellness app

🔗 Available via #GeminiAPI & #VertexAI
Integrates with #LangChain, #LlamaIndex, #Haystack, #Weaviate, #Qdrant, #ChromaDB & Vector Search
🧪 Colab notebooks for Gemini API & Vertex AI ready to use

🌐
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2-generally-available/

→ Mindlid: 20% lift in top-1 recall for wellness app

🔗 Available via #GeminiAPI & #VertexAI
Integrates with #LangChain, #LlamaIndex, #Haystack, #Weaviate, #Qdrant, #ChromaDB & Vector Search
🧪 Colab notebooks for Gemini API & Vertex AI ready to use

🌐
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2-generally-available/

→ Mindlid: 20% lift in top-1 recall for wellness app

🔗 Available via #GeminiAPI & #VertexAI
Integrates with #LangChain, #LlamaIndex, #Haystack, #Weaviate, #Qdrant, #ChromaDB & Vector Search
🧪 Colab notebooks for Gemini API & Vertex AI ready to use

🌐
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-embedding-2-generally-available/

We were looking for a local tokenizer for counting the number of input tokens before calling the gemini-embedding-001 endpoint on vertex AI. Turns out this Gemma tokenizer returns exactly the same number of tokens as the usage in the embeddings result `embedding.statistics.token_count` of the Gemini embeddings endpoint. Tested on 2000 datapoints. 😁

https://github.com/google/gemma_pytorch/blob/33b652c465537c6158f9a472ea5700e5e770ad3f/tokenizer/tokenizer.model

#gemini #embeddings #gemma #vertexai #genai

We were looking for a local tokenizer for counting the number of input tokens before calling the gemini-embedding-001 endpoint on vertex AI. Turns out this Gemma tokenizer returns exactly the same number of tokens as the usage in the embeddings result `embedding.statistics.token_count` of the Gemini embeddings endpoint. Tested on 2000 datapoints. 😁

https://github.com/google/gemma_pytorch/blob/33b652c465537c6158f9a472ea5700e5e770ad3f/tokenizer/tokenizer.model

#gemini #embeddings #gemma #vertexai #genai

We were looking for a local tokenizer for counting the number of input tokens before calling the gemini-embedding-001 endpoint on vertex AI. Turns out this Gemma tokenizer returns exactly the same number of tokens as the usage in the embeddings result `embedding.statistics.token_count` of the Gemini embeddings endpoint. Tested on 2000 datapoints. 😁

https://github.com/google/gemma_pytorch/blob/33b652c465537c6158f9a472ea5700e5e770ad3f/tokenizer/tokenizer.model

#gemini #embeddings #gemma #vertexai #genai

We were looking for a local tokenizer for counting the number of input tokens before calling the gemini-embedding-001 endpoint on vertex AI. Turns out this Gemma tokenizer returns exactly the same number of tokens as the usage in the embeddings result `embedding.statistics.token_count` of the Gemini embeddings endpoint. Tested on 2000 datapoints. 😁

https://github.com/google/gemma_pytorch/blob/33b652c465537c6158f9a472ea5700e5e770ad3f/tokenizer/tokenizer.model

#gemini #embeddings #gemma #vertexai #genai

We were looking for a local tokenizer for counting the number of input tokens before calling the gemini-embedding-001 endpoint on vertex AI. Turns out this Gemma tokenizer returns exactly the same number of tokens as the usage in the embeddings result `embedding.statistics.token_count` of the Gemini embeddings endpoint. Tested on 2000 datapoints. 😁

https://github.com/google/gemma_pytorch/blob/33b652c465537c6158f9a472ea5700e5e770ad3f/tokenizer/tokenizer.model

#gemini #embeddings #gemma #vertexai #genai

Discourse on LLM Frameworks Reveals Inconsistencies in AI Efficacy

AI experts at the Sydney LLM-UM workshop on April 29 discussed how AI models like LLMs need good data to work well and face computer limits.

#LLMWorkshop, #AISydney, #AIData, #AIlimits, #VertexAI

https://newsletter.tf/sydney-llm-workshop-ai-data-limits-april-29/

Discourse on LLM Frameworks Reveals Inconsistencies in AI Efficacy

AI experts at the Sydney LLM-UM workshop on April 29 discussed how AI models like LLMs need good data to work well and face computer limits.

#LLMWorkshop, #AISydney, #AIData, #AIlimits, #VertexAI

https://newsletter.tf/sydney-llm-workshop-ai-data-limits-april-29/

AI experts met in Sydney on April 29 to talk about Large Language Models. They found that AI is only as good as the data it's trained on and can be slow to use.

#LLMWorkshop, #AISydney, #AIData, #AIlimits, #VertexAI
https://newsletter.tf/sydney-llm-workshop-ai-data-limits-april-29/

AI experts met in Sydney on April 29 to talk about Large Language Models. They found that AI is only as good as the data it's trained on and can be slow to use.

#LLMWorkshop, #AISydney, #AIData, #AIlimits, #VertexAI
https://newsletter.tf/sydney-llm-workshop-ai-data-limits-april-29/

Doppelagenten in der Cloud: Sicherheitslücken in GCP Vertex AI aufgedeckt

Über eine fehlerhafte Standardkonfiguration der Berechtigungsbereiche und die Kompromittierung eines einzelnen Service-Agenten ließ sich das Berechtigungsmodell von Vertex AI systematisch ausnutzen.

https://www.all-about-security.de/doppelagenten-in-der-cloud-sicherheitsluecken-in-gcp-vertex-ai-aufgedeckt/

#cloud #google #vertexai #gcp #ki #ai #iam

Doppelagenten in der Cloud: Sicherheitslücken in GCP Vertex AI aufgedeckt

Über eine fehlerhafte Standardkonfiguration der Berechtigungsbereiche und die Kompromittierung eines einzelnen Service-Agenten ließ sich das Berechtigungsmodell von Vertex AI systematisch ausnutzen.

https://www.all-about-security.de/doppelagenten-in-der-cloud-sicherheitsluecken-in-gcp-vertex-ai-aufgedeckt/

#cloud #google #vertexai #gcp #ki #ai #iam

Doppelagenten in der Cloud: Sicherheitslücken in GCP Vertex AI aufgedeckt

Über eine fehlerhafte Standardkonfiguration der Berechtigungsbereiche und die Kompromittierung eines einzelnen Service-Agenten ließ sich das Berechtigungsmodell von Vertex AI systematisch ausnutzen.

https://www.all-about-security.de/doppelagenten-in-der-cloud-sicherheitsluecken-in-gcp-vertex-ai-aufgedeckt/

#cloud #google #vertexai #gcp #ki #ai #iam

Doppelagenten in der Cloud: Sicherheitslücken in GCP Vertex AI aufgedeckt

Über eine fehlerhafte Standardkonfiguration der Berechtigungsbereiche und die Kompromittierung eines einzelnen Service-Agenten ließ sich das Berechtigungsmodell von Vertex AI systematisch ausnutzen.

https://www.all-about-security.de/doppelagenten-in-der-cloud-sicherheitsluecken-in-gcp-vertex-ai-aufgedeckt/

#cloud #google #vertexai #gcp #ki #ai #iam

Doppelagenten in der Cloud: Sicherheitslücken in GCP Vertex AI aufgedeckt

Über eine fehlerhafte Standardkonfiguration der Berechtigungsbereiche und die Kompromittierung eines einzelnen Service-Agenten ließ sich das Berechtigungsmodell von Vertex AI systematisch ausnutzen.

https://www.all-about-security.de/doppelagenten-in-der-cloud-sicherheitsluecken-in-gcp-vertex-ai-aufgedeckt/

#cloud #google #vertexai #gcp #ki #ai #iam

【GCP】Vertex AI Flex PayGoを把握し、Gemini3系GAに備える
https://qiita.com/kkkdev/items/3672d6095ae0b8ea36f1?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Gemini #GoogleCloud #VertexAI

【GCP】Vertex AI Flex PayGoを把握し、Gemini3系GAに備える
https://qiita.com/kkkdev/items/3672d6095ae0b8ea36f1?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items

#qiita #Gemini #GoogleCloud #VertexAI

Gemini 3.1 Flash Image im Vertex AI Katalog gesichtet.

Das Modell, intern Nano Banana 2 genannt, verbindet die Geschwindigkeit der Flash-Reihe mit der Bildqualität der Pro-Modelle. Texturdetails und räumliche Logik wurden laut Leaks deutlich verbessert.

Entwickler profitieren von niedrigen API-Kosten, was das Modell für automatisierte Workflows attraktiv macht.

#Gemini #Google #VertexAI
https://www.all-ai.de/news/news26/nano-banana-2-leak

Gemini 3.1 Flash Image im Vertex AI Katalog gesichtet.

Das Modell, intern Nano Banana 2 genannt, verbindet die Geschwindigkeit der Flash-Reihe mit der Bildqualität der Pro-Modelle. Texturdetails und räumliche Logik wurden laut Leaks deutlich verbessert.

Entwickler profitieren von niedrigen API-Kosten, was das Modell für automatisierte Workflows attraktiv macht.

#Gemini #Google #VertexAI
https://www.all-ai.de/news/news26/nano-banana-2-leak

Gemini 3.1 Flash Image im Vertex AI Katalog gesichtet.

Das Modell, intern Nano Banana 2 genannt, verbindet die Geschwindigkeit der Flash-Reihe mit der Bildqualität der Pro-Modelle. Texturdetails und räumliche Logik wurden laut Leaks deutlich verbessert.

Entwickler profitieren von niedrigen API-Kosten, was das Modell für automatisierte Workflows attraktiv macht.

#Gemini #Google #VertexAI
https://www.all-ai.de/news/news26/nano-banana-2-leak

Google представив Gemini 3.1 Pro — ШІ для багатокрокового мислення, 3D і коду
# #3D #AI #AIModel #AIStudio #AndroidStudio #ARCAGI2 #BrowseComp #Gemini31Pro #GeminiCLI #GeminiEnterprise #Google #GoogleGemini #GPQADiamond #NotebookLM #SVG #SWEBenchVerified #VertexAI
https://gizchina.net/2026/02/22/gemini-3-1-pro-model-google-mirkuvannia/

Google представив Gemini 3.1 Pro — ШІ для багатокрокового мислення, 3D і коду
# #3D #AI #AIModel #AIStudio #AndroidStudio #ARCAGI2 #BrowseComp #Gemini31Pro #GeminiCLI #GeminiEnterprise #Google #GoogleGemini #GPQADiamond #NotebookLM #SVG #SWEBenchVerified #VertexAI
https://gizchina.net/2026/02/22/gemini-3-1-pro-model-google-mirkuvannia/

Aggiornamento su Claude #Sonnet 5:

I log #VertexAI alimentano speculazioni su un rilascio imminente con rumors di :
- pricing competitivo (-50% vs #Opus 4.5)
- 1M token context
- migliore SWE-bench.

Tuttavia, si tratta di screenshot non verificati da fonti ufficiali e infatti, al momento, #Anthropic non ha annunciato nulla.

#sonnet #vertexai #opus #anthropic #claude #LLM #ai

Aggiornamento su Claude #Sonnet 5:

I log #VertexAI alimentano speculazioni su un rilascio imminente con rumors di :
- pricing competitivo (-50% vs #Opus 4.5)
- 1M token context
- migliore SWE-bench.

Tuttavia, si tratta di screenshot non verificati da fonti ufficiali e infatti, al momento, #Anthropic non ha annunciato nulla.

#sonnet #vertexai #opus #anthropic #claude #LLM #ai

Aggiornamento su Claude #Sonnet 5:

I log #VertexAI alimentano speculazioni su un rilascio imminente con rumors di :
- pricing competitivo (-50% vs #Opus 4.5)
- 1M token context
- migliore SWE-bench.

Tuttavia, si tratta di screenshot non verificati da fonti ufficiali e infatti, al momento, #Anthropic non ha annunciato nulla.

#sonnet #vertexai #opus #anthropic #claude #LLM #ai

Aggiornamento su Claude #Sonnet 5:

I log #VertexAI alimentano speculazioni su un rilascio imminente con rumors di :
- pricing competitivo (-50% vs #Opus 4.5)
- 1M token context
- migliore SWE-bench.

Tuttavia, si tratta di screenshot non verificati da fonti ufficiali e infatti, al momento, #Anthropic non ha annunciato nulla.

#sonnet #vertexai #opus #anthropic #claude #LLM #ai

Aggiornamento su Claude #Sonnet 5:

I log #VertexAI alimentano speculazioni su un rilascio imminente con rumors di :
- pricing competitivo (-50% vs #Opus 4.5)
- 1M token context
- migliore SWE-bench.

Tuttavia, si tratta di screenshot non verificati da fonti ufficiali e infatti, al momento, #Anthropic non ha annunciato nulla.

#sonnet #vertexai #opus #anthropic #claude #LLM #ai

Log di errore Google Vertex AI rivelano Claude Sonnet 5 in arrivo il 3 febbraio 2026.

Nome in codice: Fennec.

Nuova modalità "Dev Team" o Swarm per elaborazione non lineare. 🔍🤖

#ai #claude #vertexai

Log di errore Google Vertex AI rivelano Claude Sonnet 5 in arrivo il 3 febbraio 2026.

Nome in codice: Fennec.

Nuova modalità "Dev Team" o Swarm per elaborazione non lineare. 🔍🤖

#ai #claude #vertexai

Log di errore Google Vertex AI rivelano Claude Sonnet 5 in arrivo il 3 febbraio 2026.

Nome in codice: Fennec.

Nuova modalità "Dev Team" o Swarm per elaborazione non lineare. 🔍🤖

#ai #claude #vertexai

Log di errore Google Vertex AI rivelano Claude Sonnet 5 in arrivo il 3 febbraio 2026.

Nome in codice: Fennec.

Nuova modalità "Dev Team" o Swarm per elaborazione non lineare. 🔍🤖

#ai #claude #vertexai

Log di errore Google Vertex AI rivelano Claude Sonnet 5 in arrivo il 3 febbraio 2026.

Nome in codice: Fennec.

Nuova modalità "Dev Team" o Swarm per elaborazione non lineare. 🔍🤖

#ai #claude #vertexai